Image source: https://venturebeat.com/wp-content/uploads/2013/04/facial-recognition-sports.png?resize=546%2C322&strip=all
For every advance in security technology, there must be an advance in ways to get around it. Face recognition software is an ingenious method of confirming identity. And its begging for balance from ingenious methods of thwarting it.
Face recognition software doesnt require a card or number; it only needs a good shot of your face. Everyone has a unique face that is very difficult to steal (unless you are Hannibal Lecter). This makes face recognition software seem almost foolproof. Well, nothing is foolproof.
Even though face recognition software isnt widely used now, it will be soon. Biometrics is gaining ground with identity theft being so rampant. The idea with face recognition software and digital fingerprint identification scanners is that there isnt anything physical to steal or memorize.
But that doesnt mean they cant be beaten. Here are some ideas on how to beat face recognition software.
How to Avoid Being Identified By Face Recognition Software
Disguising yourself from face recognition software is easy. You dont even have to look like youre in disguise. Thats because face recognition software uses only a limited number of points to identify people. It measures the distance between all of these points to identify you. If you obscure enough of them, the software wont be able to identify you.
Face recognition software focuses on the aspects of your face that arent likely to change much, meaning it excludes your hair and the outline of your face, which could change drastically if you get fat. So to get past face recognition software you need to alter things like your eyes, brow, nose, mouth, chin, and ears. No problem.
Glasses (sun or prescription) will take your eyes out. A beard will fuzz up your chin beyond recognition. Long hair or a hat will hide your ears. Any kind of facial jewelry will also help confuse face recognition software.
Of course this only works in a situation where a camera is viewing you from a distance. If you have to submit to a direct shot, you may have to remove your hat and glasses. In this case, squinting slightly can disguise your eyes. Opening your jaw and turning your head slightly can help too. This may be enough to get you by.
How to Make Face Recognition Software Think Youre Someone Else
Looking like someone else is much harder to do. First of all you need to get a straight on mug shot of the person you want to imitate. Then you have to measure the distances between all the points mentioned above. Be exact. Good face recognition software can measure within millimeters. Superimposing a high density grid over the mug shot is helpful.
The next step is to create the illusion. Its best if you can create a mask with the proper dimensions and molded to your face. I wont go into how to do that now, but you can find that info on the Internet. Ill give you a hint. It isnt as easy as it looks on MI3.
If you arent able to create good mask, you may be able to alter your own face enough to do the job. It wont be as sure because your sweating and movement can interfere with your artwork and youll probably need help.
Because face recognition software uses shapes and light and dark contrast to find points on your face you may be able to fool it with white and black makeup. Use makeup to amend the distance between your eyes and eye brows. Bite something of the proper width to show a correct chin length. Use Kleenex to stuff your nose and tape to put your ears in place. Youll have to do a lot of experimenting to get this right before you try to fool face recognition software.
And I know what youre thinking. Face recognition software can identify when youre just holding up a flat picture of the person youre trying to be. Sorry.
Well, I hope all of this helps you overcome your biometric limitations. Becoming someone else is a sticky job. Be aware that face recognition software may be coupled with some other biometric measure like fingerprint identification. Also, someday face recognition software will become more advanced and these tricks wont work. But I dont see that happening for at least 10 years.
About the Author: Mat Moniker is a writer for Innuity and a technology dabbler. To learn more about the face recognition software on the market go to Fulcrum Biometrics Website
